Optical Glass: Past and Future of a Key Enabling Material

Views

Optical glass: past and future of a key enabling material

Peter Hartmann glass is a very small business compared with other material SCHOTT AG – Advanced Optics, Hattenbergstra ß e 10, supply business or with glass sales for consumer applications. Mainz 55122, Germany However, optical glass is literally at the core of all optical systems and all technology relying on them would be useless e-mail: [email protected] without glass.

2. Making glass is meeting extreme technical Glass has been in use for more than 5000 years. Within the requirements past 130 years it became the subject of scientifi c advance- ment towards a material of precisely predictable and repro- Since the fi rst lithium glass that Otto Schott created in 1879 ducible properties. As such, it was a crucial element in many glass types have been developed (Figure 1 ). On intro- the technical revolutions of the 19th and 20th centuries. ducing additional chemical elements the range of index of However, glass development did not stop. Today, environ- refraction and dispersion combinations has been extended mental and not least economical aspects play an important largely. Fluoride and fl uorophosphates glasses, borosilicate, role in the dramatic change of glass materials and their barium- and lanthanum-containing glass types add low dis- availability. Almost 90% of glass types have vanished from persion variants for given refractive index levels [2] . catalogs in recent decades, whereas new materials are being The new materials pushed an enormous progress in glass rapidly developed. processing technology: some glass types required new melting processes, others are chemically so aggressive that they would 1. Technological vs. economic value destroy the traditionally used clay pots. Many of them have a high tendency to crystallization. With the traditional melting It was a long and tedious path to precision optics. The and casting method they would not become glasses at all. immeasurable and still ongoing technological revolution Changing to a continuous melting process, which is called enabled by precision optics started when optical glass tank melting, resulted in signifi cant cost reduction. One rea- became a technical material, which Otto Schott achieved son for this being the much better yield of good glass from the approximately 125 years ago in Jena, Germany, in close same amount of raw material and the other being the much cooperation with Ernst Abbe and Carl Zeiss. The progress higher energy effi ciency. made was strongly related to achievements in optical glass The continuous rise in requirements on the performance manufacturing such as improvement of glass quality or the of optical systems also had consequences on optical glasses. introduction of new glass types beyond the few existing Diffraction-limited systems being exceptional in the past have crown and fl int glasses with markedly different dispersion now become the rule even in consumer optics. This leads to properties [1] . extreme specifi cations for the glass. The subsequent developments of precision optics after Although it is not very diffi cult to fulfi ll such require- the improvements in glass quality demonstrated how large ments with glass parts as small as several millimeters in the demand had been. For example, the availability of high- diameter and thickness, challenges increase with large com- performance microscopes boosted medical research. They ponents overproportionally. Optical homogeneity of better × -6 strongly supported the successful fi ght against infection dis- than 1 10 across 5 mm diameter should be self-evident eases, which extended average human lifetime by approxi- for a piece of glass being called optical glass. Across 30 mm mately 20 additional years. diameter this is no longer a given in any case, for 200 mm it However, the strong demand for optical systems does not is an art, and beyond 500 mm it is an outstanding achieve- refl ect immediately in equally strong demand for optical glass ment. In particular, the application of optical glass in micro- material. Glass has a typical share of 1– 2 % or less of the total lithography posed extreme requirements to all specifi cation value of optical systems. Schott has not become a large com- characteristics. Even the narrowest catalog tolerances were pany on the sales of optical glass but due to sales of spin-off no longer suffi cient and had to be surpassed considerably products such as gas lamp cylinders. The borosilicate glass for all characteristics at the same time and reliably for a used in these cylinders was invented in search of new optical large series of large discs (Figure 2 ). The necessary devel- glass types. From a purely economic view the sale of optical opments in melting technology, measurement methods and quality assurance have been costly but led to sustained suc- www.degruyter.com/aot cess. Optical homogeneity levels well below 1 × 10 -6 could 6 P. Har tmann

Refractive index 2.00 Abbe diagram

N-glasses LASF KZFS-glasses 1.90 Classical glasses CaF2 and fused silica SF P-glasses 'low Tg' 1.80 Ernst Abbe 2009 LAK LAF 1.70 BASF SSK BAF SK F 1.60 PSK BALF LF BAK LLF KF PK K BK 1.50

CAF2 FK Abbe number FS Otto Schott 1.40 100 90 80 70 60 50 40 30 20

Figure 1 The Abbe diagram provides an overview for the portfolio of existing glass types plotting their optical positions within the refractive index and dispersion grid. The dispersion is characterized with the Abbe number, which is given traditionally in reverse order as dispersion rises with decreasing Abbe number.

FK5 Ø 210×38.3 PV: 4×10-7 (H6)

Wave front Wave front 1st derivative Date: 23.09.1996 Homogeneity Δn: +2.12* 10-7 (H5) Glass type: FK5 Focus C3: 0.29* 10-7 Melt no.: 116109 Astigmatism C4, C5: 0.95* 10-7 Disc no.: 10 Coma C6, C7: 0.27* 10-7 Annealing no.: 28042 Sph. aberration C8: 0.46* 10-7 Max. derivative cm-1: 0.79* 10-7

Figure 2 Optical glass blanks for use as lenses in microlithography are specifi ed with extremely narrow tolerances. With simultaneous development of melting and measurement technologies it became possible to deliver thousands of large blanks with extreme homogeneity. be reached for diameters larger than 200 mm and thickness from Europe and North America to Asia. The company Hoya of 40 mm not only for single pieces but for thousands of of Japan introduced a process for cost-effective series pro- them [3] . duction of aspherical lenses, the so-called precision molding. From preforms, which already provide highly smooth surfaces, they press fi nished lenses in precision molds. The 3. Some recent developments lenses only have to be centered and coated to be ready for mounting. The process is performed at the lowest possible Approximately 20 years ago new developments in opti- temperature just enabling plastic deformation. This method cal glass arose from Asia. First, they were underestimated preserves surface quality mostly and avoids early wear of the by the Western world probably because their main use was pressing tools. The main drawback is its limitation to small for consumer optics, a fi eld of optics that has largely moved lens diameters and thicknesses. Larger lenses require long Optical glass 7

cooling times to achieve the necessary high homogeneity and the Schott glass portfolio consolidated in subsequent years low stress birefringence. Production cycles would be too long until a sharp drop occurred close to the turn of the millen- to be economic. As the precision molds used are very expen- nium (Figure 3 ). Two trends met at that time: the demand for sive the method is only economic for large series, which are the ecoglass types and the need to abandon long-term uneco- typical of consumer optics. nomic glass types. This led to the loss of 90 % of the classi- The second trend relates to environment-friendly glass cal glass types of the Schott catalog of 1992 with respect to compositions. The portfolios of all glass manufacturers cur- the 1999 edition. Only 20 glass types remained and 66 new rently comprise large numbers of lead- and arsenic-free glass ones entered as lead- and arsenic-free replacement glass types types. The benefi t for the environment will be very moderate on important positions in the Abbe diagram. This was a huge considering the comparatively small amount of optical glass effort for Schott and for all affected industrial optics compa- produced worldwide and the extremely low risk that any of nies. Other Western glass manufacturers have not joined the its constituents may become bioavailable. Nevertheless, it effort and have abandoned the market completely or shrank is understandable that producers of consumer optics avoid to insignifi cance. lead or arsenic if they want to maintain their environmental- Since 2000, Schott introduced more than 30 new glass friendly image. Changing a large part of consumer optics types, 15 of them in the years from 2009 to 2011. The level to ecoglass types did not lead to performance impairment. now reached at slightly above 100 seems to be suffi cient for A smaller market, however, suffers from the considerably most of today’ s optical systems and restricted enough to be lower transmission of the lead-free glass types in the blue provided by a single glass manufacturer sustainably [4] . violet spectral range, this is digital projection. A true color projection needs three color channels with equal intensity. 5. Progressive and adverse trends The blue channel usually is the weakest already coming from the lamp. This and additional losses in the glass must After the remarkable changes in the product variety, a number be compensated by damping the other two channels lead- of new trends and some more traditional demands became ing to diffi cult heat management in the projector and high apparent: energy waste. In industrial optics, a lot of essential applications depend on high blue-violet transmission and those • High refractive index glass types for short system length applications become impossible with lead- and arsenic-free and for lower spherical aberration. glass types. • Low dispersion glass types for correction of secondary color aberrations. • Quality variants of glass types with high transmission in 4. A 90% change in the glass type availability the blue-violet light. • Glass types for precision molding with lowest possible The 1990s saw dramatic changes in the glass programs of transformation temperature and low interaction with the all manufacturers. Although having grown steadily since the mold material. time of Otto Schott up to a peak of 273 glass types in 1967, • Lead- and arsenic-free glass types.

Glass types in SCHOTT catalogs 300

273 252 250

202 200 203

154 150 112 104 100 96

Number of glass types 1886 101 82 86 67 50 44 28 25 20 20 20 20 0 1850 1870 1890 1910 1930 1950 1970 1990 2010 Year

Figure 3 The total number of optical glass types in the Schott catalog editions shows a dramatic change in recent history. The drop from 1992 to 1998 was due to the necessity to abandon uneconomic glass types with very low market demand and to change the glass portfolio to lead- and arsenic-free glass types. (Red line: classical glass types, black line: total number of glass types.) 8 P. Har tmann

Considering the strongly grown number of digital cameras 6. RoHS: threats beyond the ban of lead and and cameras in mobile phones one might think that the need arsenic for optical glass grows accordingly. This is not the case due to several reasons. Since 2003, another adverse trend appeared that might become a serious threat to total optical industry. In that year, • Because of its lower density and price plastic optics (or the European Commission released directives to increase the polymer optics) replace glass. This has reached a share of share of recycling of electric and electronic waste. To sup- more than 90 % with the use for spectacles. port this goal the content of several hazardous substances was • Optical elements become miniaturized, reducing the mate- limited to 0.1% , among which is lead, and to 0.01% for cad- rial needed. mium. As optical and fi lter glasses usually are components of • Different functions distributed on several elements are now electronic devices, these materials also became subject to the combined in one element. directive RoHS (Restriction of certain Hazardous Substances • Aspherical lenses replace two or more spherical lenses. 2002/95/EC) [5] . • Smaller near-net shape glass preforms (precision gobs or Together with the German Industrial Federation Spectaris rods for precision molding) give higher yield. representing many companies from optics, precision mechan- • Electronics replace optical systems as view fi nders, dis- ics and medicine and Carl Zeiss AG and with support from plays instead of pentaprisms with SLR cameras. international optical companies Schott AG succeeded in Additionally for European/US glass manufacturers. obtaining an exemption for optical and fi lter glasses in 2005. These glass types are essential for many highly important • Large glass volumes demand for consumer optics went to applications in medicine, safety, environmental surveillance Asia. and general research and development. The loss of these • European and US industrial optics manufacturers purchase applications would be in vast disproportion to any environ- material from Asia for lower prices. mental benefi t, which is so minute that it is diffi cult to prove These reasons might lead to endangering the existence its existence at all. of European and US American optical glass companies The company consulting the European Union (EU) in (Figure 4 ). Smaller demand leads to smaller production lots preparation of the exemption decision agreed with this view. and thus higher costs. A growing number of optical glass Unfortunately, the EU granted the exemption only for a types might fall below technical or economical limits and period of 4 years as a routine measure to maintain innova- hence be lost from the glass portfolio. Only a minimum profi t tion pressure on industry. In September 2010, the exemption enables developments. Reduced margins lead to fewer inno- was extended for another 4 years and by July 1, 2011 it was vations. This also holds for customer service such as provi- extended again on the occasion of the release of the revised sion of special quality grades, data and information, technical RoHS (directive 2011/65/EU, the so-called RoHS recast) to consulting, standardization work and lobbying activities. July 20, 2016. Customers expect all this from glass manufacturers but the From today ’ s point of view this might seem to be comfort- fi nancial requirements of such services are easily overlooked. ably far away in the future. However, this may turn out to be Purchasing only from the cheapest source may turn out as short-sighted. Optical designs of high-performance systems extremely expensive in the long term. need long-term reliability in optical glass availability. From The change in the glass supply landscape (see Figure 4) the start of a new design based on existing glass types usu- leaves SCHOTT AG as the sole manufacturer of optical glasses ally 2 years and more pass before the market entry of the new in the Western world. This might be already seen as a warning optical system. Even after 5–10 years market presence spare sign for a critical change in the availability of optical glasses. parts have to be guaranteed, for special systems even up to 30 years. The expiration and extension procedures established by the revised RoHS do not help to remove uncertainties from the exemption periods. The remaining time for glass types will shrink to 6 months before the expiry date, and in special cases even down to 3 months. No one can use a material for a long-term design if there is a risk that it will be prohibited before the product is launched. There is another risk for the long-term availability of optical materials. More chemical elements might be added to the prohibition list. In the revision phase of RoHS the elements arsenic and antimony were under consideration. Both elements are in use with optical glass with contents fairly below 1% , but in many cases above the usually set limit value of 0.1% . Their purpose is to prevent high levels of tiny bubbles Figure 4 The present-day world map of optical glass manufactur- in glass. This refi ning effect comes from their change in ers shows a concentration in Asia with only a small portion left in valence bonds from three at high temperature to fi ve at low Europe and in the USA. temperature, which enables them to integrate gas atoms into Optical glass 9

the atomic network of glass. Even though arsenic had already decision 6 months before expiry. The resulting research and been removed from many glass compositions it is still nec- administration effort would mean not only the complete loss essary for special applications, where classical glasses can- of profi t for many glass types, it would even be higher than not be replaced because of their unique combination of high the total turnover for many glass types and hence would lead refractive index and high blue-violet transmission. This com- to the complete loss of many glass types. With such an unreli- bination requires the presence of lead and arsenic together. able material basis precision optics would be in question as a Antimony had been the replacement element for arsenic in whole. Knowing about the key enabling character of preci- many glass types during the redevelopment of optical glasses sion optics for many other far bigger technology branches this in the 1990s. There is no replacement for antimony, so that could lead to widespread stagnation and regress. banning this element would lead to the loss of many glass types (Figure 5 ), even though some might be rescued by reducing the content below 0.1 % . 7. Exempt optical glasses from RoHS The directive RoHS asks for regular revision of the substances prohibition list if additional substances should be By contrast, any environmental benefi t is doubtful. The added. Beyond arsenic and antimony, boron, selenium, cobalt reduction potential for each possibly declared hazardous and nickel had been in discussion already. Boron would lead substance in optical glass is below one part per million in to additional losses of important glass types. With selenium, most cases, even much lower than that in relation to the total cobalt and nickel being prohibited, not a single colored fi lter amount of electric and electronic waste. Here is one exam- glass would exist any longer. ple. The total amount of cadmium used in fi lter glass was 0.3 Generally, heavy metals are in the focus of observation. tons worldwide in 2007, the amount of electrowaste just in As there is a considerable number of heavy metal elements in Europe was approximately 10 million tons. This content is use with optical glasses, fi lter glasses, special optical glasses completely negligible even if cadmium in electrowaste were and optical glass ceramic, there is a high probability that such to be bioavailable, i.e., could enter biological organisms. glass types might become subject to future prohibitions. However, it is like all other ingredients fi rmly bound in the For each element added to the prohibition list there would atomic network of the glass. There are no mechanisms allow- be the need for an exemption application. Case-by-case the ing elements to leave the bulk glass in considerable amounts, necessity of each element would have to be proven on the not even in waste incineration plants. In fact, one of the best basis of objective retrievable data considering possible alter- ways to take elements out of bioavailability is to melt them natives, the whole product lifetime and possibly exchange into glass. plans. At best, one could get 5 years exemption with the If environmental valuation is not restricted to the pres- need to apply for an extension latest 18 months and granting ence or absence of hazardous substances but widened to take

Refractive index 2.00 Abbe diagram LASF N-glasses 1.90 KZFS-glasses Classical glasses CaF2 and fused silica SF P-glasses 1.80 2009 LAK LAF 1.70 BASF SSK BAF SK F 1.60 PSK BALF LF BAK LLF KF PK K BK 1.50

CAF2 FK Abbe number FS 1.40 100 90 80 70 60 50 40 30 20

Figure 5 If the EU directive RoHS were to restrict the use of arsenic and antimony to contents below 0.1 % most optical glass types would be lost. There would be wide regions where no glasses would be left at all. Such a restriction would have severe implications for precision optics and its downstream applications. 10 P. Har tmann

energy consumption into account the lead-free glasses lose regulations are ranked as laws enforced with the possibility of their benefi ts, if there were any at all considering the argu- personal punishment. ments above. The more lead replaced in the glass composition For this reason, the next target will be to have optical materials the higher the differences in processing temperatures relevant taken out of the RoHS scope in total. High-quality optical glass, for melting, pressing and fi ne annealing. Usually, lead-free fi lter glass and optical glass ceramics are key enabling materials glass types require much more energy not only in the melting of the present and the future. Their availability is not self-evident. process but also in the subsequent reheat press processes. The users should support the supplier, who takes the effort to get The EU as well as the German government have recog- the removal done by the EU, by contributing additional cases to nized their large heritage in optics and photonics with many the collection of applications, where possibly endangered glass companies and institutions forming a broad basis of leading types are indispensable, and by open confession to this need. edge capabilities. They have realized the tremendous leverage effect of photonics on other technologies of even much higher economic importance. Therefore, they have committed them- References selves to support research and development with a high level of funding within the next 10 years. However, it is not only [1] C. R. Kurkjian and W. R. Prindle, J. Am. Ceram. Soc. 81, important to acquire new capabilities but also to maintain the 795 – 813 (1998). [2] P. Hartmann, R. Jedamzik, S. Reichel and B. Schreder, Appl. high standard, which has been achieved already and that is Opt. 49, D157 – D176 (2010) . needed for many high-ranked EU objectives such as technical [3] P. Hartmann, H. F. Morian and R. Jedamzik, in ‘ Large Lenses competitiveness, safety, health, climate and environment pro- and Prisms ’ , Eds. by R. G. Bingham and D. D. Walker (Proc. tection. The disproportion between the negative consequences SPIE 4411, Bellingham, WA, USA, 2002) pp. 6 – 20. of prohibiting chemical elements for use with optical materi- [4] SCHOTT catalog ‘ Optical Glass ’ 10.349 08112.5 (2011) . als and the extremely minute benefi t for the environment is [5] P. Hartmann and U. Hamm, Proc. SPIE 8065, 806511 (Bellingham, so extraordinary large that sometimes it is forgotten that the WA, USA, 2011) .

Peter Hartmann is Director market and customer rela- tions at SCHOTT AG, Mainz, Germany. He graduated in physics with a doctoral the- sis on the development of high energy photon scintilla- tion glasses at Max-Planck- Institute and University of Mainz in 1983. He joined the SCHOTT Optics Division in 1985 and worked in the ﬁ eld of quality assurance and product application of optical glasses and zero- expansion glass ceramic Zerodur. He was involved in a number of major telescope projects (CHANDRA, ESO-VLT, KECK, GRANTECAN, AEOS Maui, TNG Padua, VISTA, SST) and the development of i-line glass for Microlithography. Peter Hartmann serves as member of the Board of Directors of SPIE from 2011–2013.